A plurality of proposals have been made about a mechanism that attains high-speed processing by distributing a great deal of processing to a plurality of devices upon processing. For example, refer to IBM professional paper “Application of Grid Computing to Commercial Systems” for grid computing and the like.
In some cases, processing that requires huge amounts of time by fully utilizing a plurality of devices connected to a network and receiving a job from a client is executed. In such case, it is expected to implement higher-speed distributed processing by using the framework of grid computing to effectively utilize the processing capacities of a plurality of devices.
On the other hand, in recent years, the awareness of energy saving has heightened, and it is a common practice to automatically shift an unused device to a power-saving mode and to set it in a standby state. Such devices are connected via a network, and are controlled to process jobs requested from other devices. In such case, other devices request jobs regardless of the shift state to the power-saving mode and the consumption power of the job request destination device. As a result, the request destination device which will shift to the power-saving mode unless it receives job requests from other devices cannot shift to the power-saving mode, and loses chances to stand by in the power-saving mode. As a result, the energy saving effect is impaired.